W3C home > Mailing lists > Public > public-awwsw@w3.org > May 2010

Re: [pedantic-web] Re: The OWL Ontology URI

From: David Booth <david@dbooth.org>
Date: Wed, 12 May 2010 21:56:41 -0400
To: Dan Connolly <connolly@w3.org>
Cc: Pat Hayes <phayes@ihmc.us>, Jonathan Rees <jar@creativecommons.org>, AWWSW TF <public-awwsw@w3.org>
Message-ID: <1273715801.3264.25078.camel@dbooth-laptop>
On Wed, 2010-05-12 at 10:31 -0500, Dan Connolly wrote:
[ . . . ]
> "Other things, such as cars and dogs (and, if you've printed this document on physical
> sheets of paper, the artifact that you are holding in your hand), are resources too.
> They are not information resources."
> so w:InformationResource is disjoint with cars, dogs, and pieces of paper.
> That suggests fairly strongly that it's disjoint with people, but doesn't
> say so exactly. How about Graphs? Integers? I don't find any compelling
> argument based solely on ratified specs and/or decisions of the TAG.

It is pointless to make claims about what is or isn't disjoint with
w:InformationResource when we do not have either a formal definition --
a set of assertions -- of w:InformationResource or of "dog", "person",
etc.  If we make up RDF definitions of those terms *then* we can talk
about what is disjoint and what isn't.

Furthermore, as I've pointed out before, there is no architectural
*need* to define w:InformationResource as disjoint with something like
sumo:Person (from the SUMO upper ontology).  Ambiguity of resource
identity is inescapable -- we just have to accept that fact and learn to
deal with it -- so although the AWWW (correctly, in my view) admonishes
people to mint different URIs for different things, 
the choice of what to consider *different* (versus what to benignly
conflate) will always be an application-dependent judgement call that
seeks to balance immediate simplicity against support for other -- more
discerning -- uses of the URI(s).  A single URI for X may be perfectly
adequate and unambiguous for one application, while a different
application may need to distinguish between X1 and X2, which were
conflated by using a single URI for X.  

There is no significant benefit to the web architecture by trying to
insist that w:InformationResource must be distinct from some notion of
person.  The salient architectural advice was already given the AWWW
section on avoiding URI collisions:
Given that advice, it is up to URI owners to decide where to draw the
line between simplicity and support for other uses of a URI.

(Incidentally, this does not mean that the notion of "information
resource" is useless: it is an important role in the web architecture.
An "information resource" is the kind of thing that can have

I know of no other reasonable and general way to define the identity of
a resource than by specifying a set of assertions that constrain it.
And of course, in the semantic web world, the natural way to do that is
with a set of RDF assertions.  Given that it would be (IMO) completely
nonsensical for different people to arbitrarily assume completely
different resource definitions for a given URI, the next logical
question is: how should the resource identity for a URI be determined?
Or another way to state this is: how should the resource identity for a
URI at least be *minimally* constrained, given that it will be further
constrained in different applications.  This leads to the notion of a
URI declaration.  

One can think of a URI declaration as providing a set of constraints
that (minimally) limit the possible interpretations for that URI that
are permissible.  Of course, when a URI is used in a semantic web
application there will normally be *additional* assertions involving
that URI, and those will additionally constrain the potential
interpretations for that URI, just as illustrated in step 3a of figure

Thus, it is perfectly reasonable that one URI owner may mint a URI that
denotes something that is both an w:InformationResource and an RDF
graph, while another URI owner chooses to distinguish between them by
minting separate URIs.  Again, without a definition of the resource
identity for each URI (e.g., a URI declaration for each) there is no
basis for deciding whether they are disjoint or not.

David Booth, Ph.D.
Cleveland Clinic (contractor)

Opinions expressed herein are those of the author and do not necessarily
reflect those of Cleveland Clinic.
Received on Thursday, 13 May 2010 01:57:09 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:21:08 UTC